Re: Initial window placement

[Owen Taylor <otaylor redhat com>]

"Valeriy E. Ushakov" <uwe ptc spbu ru> writes:

> On Mon, Aug 13, 2001 at 14:19:56 -0500, Sasha_Vasko osca state mo us wrote:
> 
> > > | when calculating reference point at the time of initial placement -
> > > | initial window's size should be taken into consideration, as if it
> > > | was the frame for this window.
> > >
> > > Well to me this "clarification" seems even more confusing than the
> > > original.
> > >
> > > ICCCM says that window is "SHIFTED to make room for the window manager
> > > frame".  I definitely read this as that the window must not be resized
> > > and the frame is added around the window.
> > >
> > > While ICCCM distinguishes window refpoint vs frame refpoint, the
> > > "clarification" speaks about some unqualified "reference point" and
> > > lumps both window size and frame size in one phrase in a very
> > > confusing manner.  So why this "clarification" is necessary in the
> > > first place?
> > 
> > Well, ICCCM description of all that is garbadge and over years
> > generated nothing but confusion.
> > 
> > While reading the entire paragraph you'll find out exactly what
> > "reference point" specs are talking about, how it should be
> > calculated and how it should be applied. Clear formulas. No
> > confusion of frame/client ref points. Just copy and paste it in your
> > code - and you are done.
> 
> [Come on.  It's really tempting to return that "garbage" argument, but
> it will be obviously unproductive.]
> 
> If you think that ICCCM wording is bad and confusing, than it doesn't
> help to switch to a different terminology to clarify it without
> providing any reference points (pun intended).  Whether you like it or
> not, but ICCCM talks about frame/client ref points, so denial will
> just do no good.  If that language is confusing than take trouble to
> explain in the text why it's confusing, what new language you propose
> and how they are related to each other.

There is a clear reason for adding it the note in the new window spec:
almost no window managers get the case of XMoveWindow right.

It is an attempt to rephrase the positioning algorithm in a
way that doesn't talk about "shifting" an original position,
but simply about where the window manager frame is positioned
given the location specified by the client.

(I find the wording a little confusing ... it needs to be read
over a few times to get the meaning. But it is unambiguous.
Which can't be said for the description in the ICCCM .

Regards,
                                        Owen